What are variants in A/B testing?

• A. The different versions of content served to users, with Variant A being the original and Variant B being a modified version
• B. A single version of content tested on all users
• C. A measurement metric like click-through rate
• D. The schedule for running the test

Answer: (A)

Variants in A/B testing are the different versions of content shown to users. The idea is to compare how each version performs on a chosen outcome, such as click-through rate or conversions. Typically, Variant A is the original version and Variant B is a modified version. By exposing distinct groups of users to each version, you can see which one drives better results and make data-driven decisions.

A single version isn’t a variant because there’s nothing to compare. A measurement metric isn’t a version of content either—it’s what you evaluate after running the test. And the schedule refers to when you run the test, not the actual versions being tested.